On Wed, May 08, 2013 at 10:03:21AM -0400, Chris Lumens wrote: > > I'm probaby doing something really silly, or missing something obvious. I'd > > appreciate if someone could help me figure out where! > Your pykickstart bit looks fine, and you can prove that by running "make > test" or by running ksvalidator against a file containing your new flag. > So it's gotta be something in the integration. Yeah, did the former, and now with ksvalidator too (a useful tool to know about in its own right!). [...] > Your anaconda patch works by modifying get_bootloader to set the > bootloader if a command line argument has been provided. That happens > really early on. At the point where bootloader.execute runs, it's too > late. Okay, yeah, obvious in retrospect. > What might be easiest is if you modify dracut/parse-kickstart instead of > pyanaconda/kickstart.py. Put in a block for bootloader similar to > what's already there, and have it add the command line argument if the > flag is given in kickstart. That'd save you from having to write code > to change out the anaconda.bootloader instance. *nod* I assume the pykickstart patch should also be there so the validator works and so other tools can use it? > Who knows what'll go wrong with that. There are a lot of moving parts here. :) > Just keep in mind that once this option is in, we'll have to support it > for a good long while. Yep. I'm committed. :) (I stuck my name at the top of the file with the earlier extlinux patches, so people know who to find if something goes wrong.) > > - Chris > > _______________________________________________ > Anaconda-devel-list mailing list > Anaconda-devel-list@xxxxxxxxxx > https://www.redhat.com/mailman/listinfo/anaconda-devel-list > -- Matthew Miller ☁☁☁ Fedora Cloud Architect ☁☁☁ <mattdm@xxxxxxxxxxxxxxxxx> _______________________________________________ Anaconda-devel-list mailing list Anaconda-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/anaconda-devel-list
